Change unassigned filter

[Lunga001]

Description of the issue/feature this PR addresses

On the AR Listing change the Unassigned filter because it currently displays a cluttered list

Current behavior before PR

Filters ARs that are not assigned to a Worksheet by SampleReceived, ToBeVerified, AttachmentDue, Verified and Published states

Desired behavior after PR is merged

Should filter ARs that are not assigned to a Worksheet by SampleRegistered, SamplePrep, ToBeSampled, SampleDue, SampleReceived, and AttachmentDue states

--
I confirm I have tested this PR thoroughly and coded it according to PEP8
and Plone's Python styleguide standards.

[xispa]

With this Pull Request, when the Unassigned filter from Analysis Requests lists is used, only Analysis Requests that meet the following criteria are displayed:

• The Analysis Request hasn't been yet transitioned to to_be_verified state (this is, the Analysis Request has at least one result that hasn't been yet submitted).
• At least one of the Analysis associated to the Analysis Request has not been assigned to a Worksheet.

Note that the assigned/unassigned status is only meaningful during the planning of the tasks to be done by analysts, when there are no results set yet. In fact, assigned status means that all the analyses from a given AR has been assigned to a Worksheet and, therefore to an analyst. Thus, ARs that are in a state beyond received (to_be_verified, verified, published, ...) are not displayed.

[xispa]

sample_registered is a temporary state the Analysis Request reaches when is created, but the object is immediately transitioned to either sampled (if sampling workflow disabled) or to_be_sampled (if sampling workflow is enabled). In turn, if no preservation is required, an Analysis Request with sampled state is automatically transitioned to sample_due. In summary, querying for sample_registered should never return anything.

From my point of view, these are the states to be filtered by when unassigned filter is used: sample_received, sample_due, sample_prep, to_be_preserved, to_be_sampled, attachment_due.

Also filter by cancellation_state == 'active' (we don't want inactive ARs to be listed)

[xispa]

My wrong. The assignment of analyses to Worksheets can only be done for those analyses that belong to Analysis Requests that are associated to a Sample Partitions that have been received. Therefore, there is no sense to add in this filter those states that are from the preservation or sampling workflow. Following this reasoning, I suggest this filter to consider the following states:

{'worksheetanalysis_review_state': 'unassigned',
 'cancellation_state': 'active',
 'review_state': ('sample_received', 'attachment_due')

Related, but for analyses instead of Analysis Requests: Analyses filter not kept after selecting one in dashboard #613 . Note the comments:

• Analyses filter not kept after selecting one in dashboard #613 (comment)
• Analyses filter not kept after selecting one in dashboard #613 (comment)

[xispa]

Transitions publish, republish, verify and reinstate will never be available for ARs displayed when using this unassigned filter: only active (so reinstate will never be used) and those ARs that at least have one Analysis with result pending will be displayed (so verify, publish and republish will never be used).

[Lunga001]

Thank you for review @xispa, I've made the requested changes using these filters
{'worksheetanalysis_review_state': 'unassigned',
'cancellation_state': 'active',
'review_state': ('sample_received', 'attachment_due') }

[xispa]

I realized now that this filter will never work because of these two reasons:

• Analysis Requests are never transitioned to assigned/unassigned #637
• The AnalysisRequests catalog does not have an index for worksheetanalysis_review_state

I've fixed the first, would be nice @Lunga001 if you fix the second before I accept this PR. Remember to add the necessary logic into the upgradestep.

EDIT: both points mentioned above addressed in #637